Implied probability turns betting odds into a percentage. At decimal odds of 2.50, the implied chance is 40%: 1 ÷ 2.50 × 100 = 40. That 40% is the win rate needed to break even at the quoted price. It is not proof that the outcome has a 40% chance of happening.
Decimal odds are the familiar format for British sports and racing. Converting them is easy; knowing what to do with the percentage is harder. A bookmaker may include a margin in its price, so a market-derived probability and an independent estimate of the event's chance can differ.

What does implied probability mean?
Shorter decimal odds imply a higher chance but pay less profit per dollar staked. Longer prices imply a lower chance and pay more if the selection wins. The calculation describes the price for a particular selection, not the strength of a team or runner by itself.
Imagine 100 identical £10 bets at 2.50, each settled as a full win or loss. Forty wins return 40 × £25 = £1,000, exactly matching the £1,000 staked. With 39 wins the bettor loses money; with 41, the bettor makes a profit. The example assumes no fees, voids or price changes.
How to calculate implied probability from decimal odds
Implied probability (%) = (1 ÷ decimal odds) × 100
At 1.72, the calculation is 1 ÷ 1.72 × 100 = 58.14%. At 4.00 it is 25%. Even-money odds of 2.00 correspond to 50%.
Stake, total return and profit are different numbers
A £20 win bet at 2.50 returns £50 if it wins: £20 of the original stake and £30 profit. The price still implies 40% whether the stake is £2, £20 or £200.
Total return = stake × decimal odds
Net profit on a win = stake × (decimal odds − 1)
Implied probability chart for common decimal odds
| Decimal odds | Implied probability | Decimal odds | Implied probability |
|---|---|---|---|
| 1.20 | 83.33% | 2.00 | 50.00% |
| 1.40 | 71.43% | 2.20 | 45.45% |
| 1.50 | 66.67% | 2.50 | 40.00% |
| 1.72 | 58.14% | 3.00 | 33.33% |
| 1.80 | 55.56% | 4.00 | 25.00% |
| 1.90 | 52.63% | 5.00 | 20.00% |
| 1.95 | 51.28% | 10.00 | 10.00% |
Fractional and American odds: the same probability in another format
Fractional odds, familiar from UK racing, and American odds (moneyline odds) express prices in different ways. Use the formula for the format quoted; the underlying payout need not change.
| Odds format | Formula for implied probability (%) | Worked conversion |
|---|---|---|
Decimal D | (1 / D) × 100 | 2.50 → 40% |
Fractional a/b | [b / (a + b)] × 100 | 3/2 → 40% |
American positive +A | [100 / (A + 100)] × 100 | +150 → 40% |
American negative −A | [A / (A + 100)] × 100, with A as a positive number | −150 → 60% |
Fractional odds
At fractional odds of 3/2, a £2 stake earns £3 profit plus the stake back. That is decimal 1 + (3 / 2) = 2.50, with a 40% implied chance. Evens (1/1) correspond to decimal 2.00.
American odds
American +150 quotes a profit of 150 units on a 100-unit stake: decimal 2.50, or 40%. A price of −150 requires 150 units at risk to win 100 in profit: decimal 1.666… and 60%. The negative sign belongs to the odds notation, not to the probability.
When comparing several prices, work from the quoted odds rather than repeatedly converting rounded percentages. Each round-off can introduce a small error.
How to turn a probability into fair decimal odds
The conversion works in reverse too: fair decimal odds = 1 ÷ estimated probability, where probability is written as a number from 0 to 1.
For a 60% estimate, 1 / 0.60 = 1.666…, usually displayed as 1.67. At 25%, the fair price is 4.00. Those prices are fair relative to the supplied estimate, before charges or bookmaker margin; they cannot tell you whether the estimate is accurate.
Implied probability vs true probability vs no-vig probability
Raw implied probability The break-even percentage calculated directly from the quoted odds. It retains any margin included in the price. True probability The actual chance of the outcome under the stated conditions. It is unknown before the event and cannot be extracted with certainty from a betting price. Fair or no-vig probability A market-derived estimate after a method has removed the quoted bookmaker margin. It is useful for comparison, but the method and underlying market can both be imperfect.A quoted price need not represent a bookmaker's pure sporting forecast. The operator may adjust it for margin, customer demand, risk exposure, team news or market movement. Reverse-engineering its exact belief from the price is not possible.
Bookmaker margin, vig and overround
To measure the margin in a market, convert all mutually exclusive outcomes that cover every possible result into percentages, then add them. The total for a conventional fixed-odds book usually exceeds 100%.
Market total (%) = sum of all raw implied probabilities
Overround (percentage points) = market total (%) − 100%
Suppose both sides are offered at 1.90. Each implies 1 / 1.90 = 52.63%. Together they total 105.26%, an overround of 5.26 percentage points. The excess is often called the bookmaker's margin, vig or juice.
Overround is not the same thing as profit on turnover. At 1.90/1.90, an idealised, evenly balanced book has a payout rate of 1 / 1.05263 = 95%, equivalent to a 5% hold, not 5.26%. Real outcomes also depend on how customers stake their money, exposure and settlement.
Always check that the prices cover the same event, betting period and settlement rules. A market can total around 100%, or even less with certain promotions or cross-operator quotes, but a mixed set of prices does not give a reliable single-book overround.
How to remove the margin: a two-outcome worked example
Consider a hypothetical two-outcome head-to-head market: Outcome A at 1.80, Outcome B at 2.10. Their raw implied chances are 55.56% and 47.62%, respectively. The market total is 103.17%.
With proportional normalisation (multiplicative de-vigging), divide each raw implied chance by the market total:
No-vig probability for an outcome = that outcome's raw implied probability ÷ sum of all raw implied probabilities
| Selection | Quoted decimal odds | Raw implied chance | No-vig chance | No-vig decimal odds |
|---|---|---|---|---|
| Outcome A | 1.80 | 55.56% | 53.85% | 1.86 |
| Outcome B | 2.10 | 47.62% | 46.15% | 2.17 |
| Market total | — | 103.17% | 100.00% | — |
That method spreads the margin proportionally. Bookmakers need not load the same margin onto favourites and longshots, a complication associated with favourite–longshot bias. Power, additive and Shin methods make different assumptions, so their no-vig estimates can disagree. None establishes the true probability.

One isolated price, such as 1.80 without the opposing quote, supplies a raw implied chance but not enough information to remove the market's full margin.
Three-way markets: the draw must be included
A football 1X2 market covers a home win, draw and away win in normal time. Imagine Wellington Phoenix quoted at 2.10 to win, 3.40 for the draw and 3.80 for the visitors in an A-League match. These are illustrative prices, not live odds.
| Result | Decimal odds | Raw implied chance | Proportional no-vig chance |
|---|---|---|---|
| Home win | 2.10 | 47.62% | 46.08% |
| Draw | 3.40 | 29.41% | 28.46% |
| Away win | 3.80 | 26.32% | 25.46% |
| Total | — | 103.35% | 100.00% |
Leave out the draw and the denominator is wrong. The same issue arises with horse-racing fields and tournament outrights: every outcome has to appear in the calculation, including any 'other' selection offered.
What implied probability tells you about value and expected value
Expected value (EV) uses a separate estimate of the chance of winning to assess the return offered by the odds. For a win-or-lose bet with no fees or special settlement conditions:
Expected ROI = (estimated probability × offered decimal odds) − 1
Suppose a forecast puts the chance at 48% and the available odds are 2.20. The quoted break-even rate is 45.45%; the forecast gives (0.48 × 2.20) − 1 = +0.056, or +5.6% expected ROI. At £20 per bet, that amounts to £1.12 in modelled expected profit per comparable bet over time. It does not predict a £1.12 gain on the next result.
Now change the estimated chance to 44%, leaving the price at 2.20. Expected ROI becomes (0.44 × 2.20) − 1 = −3.2%. A few percentage points of forecasting error can reverse the answer, before commission or unusual settlement terms enter the picture.
Why beating the no-vig market probability is not sufficient
A 1.90/1.90 two-sided market has a proportional no-vig estimate of 50% per outcome. Yet a bet actually placed at 1.90 must win 52.63% of the time to break even. Even a 51% independent estimate, above the market's no-vig 50%, gives (0.51 × 1.90) − 1 = −3.1% modelled ROI at that price.
Where an independent probability comes from
An independent sporting forecast might use player availability, confirmed teams, injuries, team strength, home advantage, recent results, rest days, weather and ground conditions. Its real test is performance on events outside the training data. If predictions labelled 70% win noticeably less often across a sufficiently large, comparable sample, the model is poorly calibrated.
Where the calculation changes across betting markets
Decimal odds can be converted to implied probability in the same way for football, racing or any other event. The market rules still matter: each-way terms, exchanges, in-play prices and same-game combinations require separate treatment. Check the settlement conditions before comparing prices.
Implied probability for multis and same-game multis
A multi (accumulator) combines several legs. Multiply their quoted decimal odds: 1.80 × 2.50 = 4.50. That combined price implies a break-even chance of 1 / 4.50 = 22.22%.
You can multiply the individual probabilities to estimate the chance that every leg wins only when the outcomes are independent. An NRL team's win and one of its players scoring in the same game are connected events. Correlation changes their joint chance; a same-game multi can also carry extra margin or dependence adjustments in the quoted price.
Odds movement, in-play prices and market comparisons
If a quote shortens from 2.50 to 2.20, implied probability rises from 40.00% to 45.45%—a shift of 5.45 percentage points. Injuries, team news, weather, incoming stakes, liquidity and a bookmaker's exposure may all affect prices. The move alone does not identify its cause or prove the outcome has become more likely.
In-play odds can move within seconds as the score or clock changes; markets can also be suspended. Closing line value (CLV) compares the earlier quote with a later reference price. Beating that reference does not guarantee profit or confirm that the original probability estimate was sound.
Compare like with like: the event, selection, line, period and settlement terms must match. An odds boost or bonus bet may change how the stake is treated, what can be withdrawn and which limits apply. The highest-looking price is not necessarily the best effective return.
Exchange prices, commission and prediction markets
On a betting exchange, back and lay prices differ, and liquidity, the bid–ask spread and commission matter. With back odds of 2.10 and a hypothetical 2% commission on winning profit, the effective decimal price is 1 + (2.10 − 1) × 0.98 = 2.078. Its break-even chance is 1 / 2.078 = 48.12%, not 47.62%.
A binary prediction-market contract priced at 40 cents and paying 100 cents on success suggests roughly 40% before fees or spreads. Contract terms, liquidity and legal access differ from fixed-odds betting, so matching headline percentages need not mean matching costs.
Common errors when interpreting implied probability
- Calling a price a forecast: the raw percentage can contain a bookmaker's margin.
- De-vigging one selection: the calculation needs the complete market, including a draw or 'other' where offered.
- Using the no-vig rate as the break-even threshold: expected value must be assessed at the actual quoted price.
- Multiplying correlated legs: a same-game multi needs a joint probability, not independent-leg arithmetic.
- Forgetting costs and settlement: commission, pushes, voids and dead heats change the effective return.
- Trusting an untested model edge: an inaccurate or poorly calibrated forecast can make a positive EV estimate illusory.
Frequently asked questions
How do I calculate implied probability without a calculator?
Divide 100 by the decimal odds. At 4.00, 100 / 4 = 25%; at 5.00, it is 20%. With a less convenient price of 1.85, the answer is about 54.05%.
Why do the probabilities in a betting market add up to more than 100%?
The individual prices normally contain a bookmaker margin. Adding the implied chances across every outcome exposes that excess, known as overround; it is not another possible result of the match.
Is no-vig probability the true probability?
No. No-vig methods remove an assumed margin from the quoted market prices. They do not reveal the true chance, and different methods can disagree on favourites and outsiders.
Can the implied probability of a single selection exceed 100%?
No ordinary decimal price of 1.00 or more gives a single-selection implied chance above 100%. The total across a full market can exceed 100% when the odds include margin.
Gambling support in Great Britain
Gambling can cause harm. For information about gambling support in Great Britain, see the NHS gambling support guidance.
